A significant portion of the ethylenediamine (EDA) made commercially is by the continuous reaction of monoethanolamine (MEA) and ammonia in the presence of hydrogen over a fixed bed reductive amination catalyst. The reaction unavoidably generates a variety of polyalkylene polyamines as well. Illustrative of many of them are the following:
AEEA--N-(2-aminoethyl)ethanolamine PA0 HEP--N-(2-hydroxyethyl)piperazine PA0 DETA--Diethylenetriamine PA0 AEP--N-(2-aminoethyl)piperazine PA0 TETA--Triethylenetetramine PA0 TEPA--Tetraethylenepentamine PA0 PEHA--Pentaethylenehexamine TETA Isomers: PA0 NTEA--Nitrilotrisethylamine PA0 TETA--Triethylenetetramine PA0 DiAEP--Diaminoethylpiperazine PA0 PEEDA--Piperazinoethylethylenediamine TEPA Isomers: PA0 AETETA--4-Aminoethyltriethylenetetramine PA0 TEPA--Tetraethylenepentamine PA0 AEPEEDA--Aminoethylpiperazinoethylethylenediamine PA0 PEDETA--Piperazinoethyldiethylenetriamine PA0 "The apparently simple reaction of NH.sub.3 and MEA to yield EDA is deceptive, for several catalytic steps are involved, and a number of side reactions occur." PA0 AEEA/EDA mole ratio: 0.56 PA0 AEEA/DETA mole ratio: 3.28 PA0 EDA/PIP+AEP+HEP mole ratio: 2.53 PA0 AEEA/PIP+AEP+HEP mole ratio: 1.43
Gibson, U.S. Pat. No. 4,400,539, patented Aug. 23, 1983, describes such a process for the manufacture of ethylenediamine. The patent describes a continuous process involving the reductive amination of MEA. A further elaboration of the process of the Gibson patent is described in Winters, U.S. Pat. No. 4,404,405, patented Sept. 13, 1983. The Gibson and Winters patents characterize the relative commercial values attributed to ethylenediamine and piperazine. The direction which the art has been moving in recent years has been away from piperazine towards a more favored product, EDA.
The compositions generated by that reaction are dependent upon a number of factors such as the selection of catalyst for carrying out the reaction, the ratio of reactants, the temperature, the pressure, reactant flow velocity through the catalyst zone, the shape and form of the catalyst, the presence and absence of other reactants such as water, and the like considerations. There are a wide variety of reductive amination catalysts for this reaction. Typically, they are viewed as hydrogenation catalysts. The most prevalent of them utilize nickel as an important ingredient. Raney nickel has for years been viewed as a good hydrogenation catalyst. In recent years, nickel in combination with other metals has grown to be a favored catalyst for effecting the reductive amination reaction.
An excellent survey of that process and the alkyleneamines compositions generated thereby can be found in Barnes et al., Ind. Eng. Chem. Prod. Res. Dev. 1981, 20, 399-407. Table II, at page 402, lists a variety of patent examples and the disclosed products of MEA ammonolysis.
According to Barnes et al.
The variety of products generated by that reaction reflects the complexity of the chemistry involved.
Until recently, AEEA was viewed to be an unwanted by-product of a number of processes (including reductive amination) for the manufacture of alkyleneamines. It has long been considered to be a precursor to PIP. See U.S. Pat. No. 2,479,657 and 3,383,417. The value of AEEA has risen considerably in recent years because significant commercial uses have evolved for it. Because the commercial processes have been designed to produce EDA and minimize the formation of PIP, owing to the limited demand for PIP, little is known about the manipulation of the commercial processes to generate larger amounts of AEEA. There are a few patents directed to the manufacture of AEEA. They do not rely on the reaction of MEA and ammonia.
Lichtenwalter, U.S. Pat. No. 3,383,417, patented May 14, 1968, describes a process for the manufacture of aminoethylethanolamine (AEEA) by the reaction of monoethanolamine (MEA) in the absence of added ammonia with a catalyst containing a major amount of nickel, copper and a minor amount of chromium oxide, manganese oxide, molybdenum oxide and/or thorium oxide at a temperature of 150.degree. C. to 250.degree. C. and a pressure of 2000 to 4000 psig. correlated so as to provide for the total conversion of about 10% to 30% of the monoethanolamine. The reaction conditions include a pressure of about 2,000 to 4,000 psig., composed of at least 80% of hydrogen partial pressure. A critical factor of the invention is to limit the degree of conversion of the MEA to no more than 30%. The examples of the patent demonstrate that "only a minor amount of aminoethylethanolamine was obtained with conversion about 40%". As the MEA conversion proceeds above 30%, the AEEA yield becomes minor in the reaction product relative to the yield of piperazine (PIP). The compositions (in mole percents) made by this process are set forth in Tables 1 and 2 of the patent and are set out in the following table:
______________________________________ AEEA PIP + AEP + HEP + EDA + AEEA DETA EDA PIP AEP HEP DETA ______________________________________ 75.5 0.7 5.8 15.7 1.4 0.9 3.08 75.5 0.4 11.9 11.6 0.3 0.3 3.08 63.3 0.7 8.5 24.0 2.6 0.7 1.73 50.5 0.6 8.6 30.3 6.6 3.2 1.02 31.0 0.8 2.3 41.8 15.3 8.8 0.45 28.6 3.1 10.4 37.5 9.6 10.7 0.40 47.4 0.9 10.7 25.8 11.2 3.9 0.90 24.1 -- 8.4 24.1 8.4 3.6 0.54 47.8 1.7 6.1 31.0 8.1 5.2 0.92 45.6 1.1 8.1 28.8 8.9 7.5 0.84 29.6 -- 7.8 41.2 11.9 9.5 0.49 25.7 1.0 7.6 43.8 12.6 9.3 0.35 12.5 0.9 5.6 51.5 13.4 10.2 0.18 ______________________________________
Lichtenwalter demonstrates the effects of operating under severe conditions of the combination of high temperature and pressure. The reaction of MEA with itself involves the condensation of amino group with hydroxyl group. The production of EDA reflects the cleavage of hydroxylethyl from AEEA, an apparent deficiency in the reaction. The more EDA formed, the more inefficient is the reaction. The degree of formation of PIP, AEP and HEP demonstrates the extent that AEEA is cyclized. The condensation of MEA with itself ("intramolecular condensation") can be pushed to primarily form cyclic amines. The formation of EDA is not part of the scheme of the condensation reaction, yet in the reductive ammonolysis reaction, EDA is an unavoidable impurity.
Ford, et al., U.S. Pat. No. 4,560,798, patented Dec. 24, 1985, utilizes a process in which the reactant, MEA, is vaporized and fed to the catalyst in the vapor state. The catalyst employed is not a reductive amination catalyst as in U.S. Pat. No. 3,383,417, but rather a specific rare earth or strontium metal hydrogen phosphorus-containing compound. The reaction is carried out at 50 to 400 psig. and at a temperature of from 175.degree. C. to 275.degree. C. No hydrogen is provided in the feed. In addition, ammonia is not included as a reactant. According to the data characterized in Table 1, at Column 6 of the patent, it is quite evident that as temperature is raised the percent conversion also goes up and the selectivity to AEEA goes down materially.
A major distinction between the process of Lichtenwalter and Ford, et al. is the presence of EDA as a significant product in Lichtenwalter's process and its apparent absence from the Ford, et al. process. This results in the Ford, et al. process producing large quantities of HEDETA rather than the DETA and EDA generated by the Lichtenwalter process. Lichtenwalter fails to suggest the production of HEDETA.
Lichtenberger, et al., U.S. Pat. No. 3,068,290, patented Dec. 11, 1962, referred to in Table II of Barnes et al., supra, describes the batch manufacture of EDA by the reaction of MEA and ammonia in the absence of hydrogen in the presence of Raney nickel in an autoclave, and depicts in Example III, the following composition, exclusive of MEA and water:
______________________________________ Mole Product Weight (g.) Moles Mole % Ratio ______________________________________ EDA 260 4.48 56 PIP 70 0.81 10 DETA 103 1.0 13 AEEA 138 1.33 17 TETAS + 48 .33 4 VARIOUS AMINES AEEA/EDA .30 AEEA/DETA 1.31 ______________________________________
Because the various amines are unknown compositions, they were treated as TETAS for the purposes of this characterization.
Johansson, et al., U.S. Pat. No. 3,766,184, patented Oct. 16, 1973, criticizes the commercial applicability of the catalyst, and hence the process, of Lichtenberger, et al., see the discussion at col. 2, lines 18 to col. 3, line 5. However, a point not raised by Johansson, et al. is the absence of hydrogen in the process of Lichtenberger, et al. The absence of hydrogen makes the process of Example III of Lichtenberger, et al. not readily reproducible, if ever reproducible, in a continuous process.
It is fairly well appreciated in the reductive amination art that the reductive amination catalyst must be first reduced before effecting the reaction, and then continuously reduced during the course of the reaction in order to keep the catalyst active and functioning. The degree of this reduction will determine the catalyst's productivity and selectivity to products. In the case of Lichtenberger, et al.'s Example III, the Raney nickel had to have been reduced with hydrogen prior to use. It is that level of hydrogenation that will sustain the reaction. As the level of the catalyst's surface becomes depleted of bound hydrogen, the catalysts activity is reduced and its composition is changed. Eventually the catalyst will become depleted of its bound hydrogen, and when that occurs, the reaction terminates. But in the course of the reaction, the composition of the reaction product changes. What is found in the reactor is a composite composition that changes over the course of the limited life of the catalyst.
Moss, et al., U.S. Pat. No. 3,151,115, patented Sept. 29, 1964, describe a process for the "maximum recovery of piperazine and N-aminoethylpiperazine," obtained at a MEA conversion "of about 75 to 95%", and depict the following table in column 6:
TABLE ______________________________________ Data From U.S. Pat. No. 3,151,115 1 2 3 4.sup.1 5.sup.1 ______________________________________ Reactor Temp., .degree.C. 199 198 198 223 227 Hydrogen rate: SCFH 800 800 800 800 800 Reaction press., psig 2,800 2,800 2,800 2,800 2,800 Feed rates, gal./hr: NH.sub.3 (anhydrous) 94 70 47 80 80 Monoethanolamine (MEA) 35.5 25.5 16 158 158 Space velocity, 6.2 4.5 3.0 3.1 3.1 g./hr. ml. cat. Conversion of MEA, 29.3 39.8 53.8 71.4 75.4 percent Yields (molar), percent: Ethylene diamine (EDA) 41.3 38.5 36.2 12.0 15.0 Piperazine (PIP) 13.7 18.7 23.2 47.6 58.0 Diethylenetriamine (DETA) 7.1 7.6 9.3 9.0 3.3 N-aminoethylpiperazine (AEP) 2.1 3.0 3.0 15.0 17.5 Aminoethylethanolamine 23.3 18.4 13.2 -- -- (AEEA) Hydroxyethylpiperazine (HEP) 0.5 2.2 3.2 6.0 3.5 Residue 12.0 11.6 11.9 10.4 2.7 PIP + AEP 15.8 21.7 26.2 62.6 75.5 ______________________________________ .sup.1 Also used 140 gal. water/hr. in feed
The following mole ratios were determined from the data in column 1 of the table from Moss, et al.:
The process of the Moss, et al. patent employs high pressures and through the loss of conversion and temperature reduction, coupled with a few other variable changes, generated a relatively high yield of AEEA in a non-preferred mode of operation. However, in doing so, the patentee generated a large amount of residues (presumed to be TETAs, TEPAs and other higher polyamines), and a small amount of DETA, without a clue to how one could avoid that result. In the case of Moss, et al., there is described an example which utilizes costly high pressures combined with low conversion to produce a composition rich in AEEA and also rich in residues, and poor in DETA, without the capacity of lowering the residues content and enriching the DETA content.
It would be desirable to have continuously produced compositions generated by the self-condensation reaction of MEA over a fixed bed of a reductive amination catalyst under commercial conditions that are rich in AEEA, and that are not disproportionately high in EDA, DETA, PIP and other cyclics.
It would be desirable to have a process which increases one's ability to generate with commercially functional catalysts the manufacture of AEEA without generating large amounts of EDA, DETA and cyclic alkylenepolyamine products. In addition, it is also desirable to have a process which is dedicated to the manufacture of AEEA over a reductive amination catalyst without the need of using costly and stringent process conditions as exemplified in U.S. Pat. No. 3,383,417. It would be most desirable to have the ability to generate AEEA at low pressures such as are typically employed in the condensation of MEA over an acid condensation catalyst as exemplified in U.S. Pat. No. 4,560,798, but to be able to use instead a reductive amination catalyst. These features are provided by the process of this invention.